1. Field of the Invention
The present invention relates to a technique which performs a process for recommending a user terminal device other documents relevant to a browsed document that is browsed by the user terminal device. More particularly, it relates to a technique which provides in addition to recommended document information, the recommendation histories of the recommended documents.
2. Background Art
Using a user terminal device such as desktop computer or mobile computer, it is performed to browse a document stored in the memory of the user terminal device or a document acquired through a network, by displaying the document on a screen.
In such a document browsing process, it is performed that a user acquires the document to-be-browsed by retrieving the document from the memory, a data server, a Web site or the like.
A typical information retrieval method in the related art is such that, in response to the input of a retrieval keyword as is explicitly performed by the user, documents in which the keyword is contained or document information items which are endowed with the keyword as an index are outputted as retrieved results.
With such an information retrieval method, however, the user must give a retrieval instruction by creating a retrieval expression which uses the retrieval keyword explicitly, and he/she must seek the necessary document from among the obtained retrieval results.
On the other hand, notice is taken of an information recommendation technique in which relevant document information items are retrieved automatically on the basis of specified document information and without the creation of any retrieval expression by a user, so as to present the retrieved information items to the user.
In recent years, the information recommendation technique has been noticed for its utilization in Internet shopping (the recommendations of relevant commodities based on a purchase history) or a news broadcast service (the provision of past news relevant to a news article). Besides, the information recommendation technique is being noticed for the active use of a corporate memory or an expansion to a knowledge management, owing to cooperation with a document archive system.
There has been known, for example, a system wherein a feature word is extracted from the content text of a Web page being browsed by a Web browser, documents of relevant contents are automatically retrieved from a specified database by using the feature word, and the retrieved documents are provided to a user terminal device (refer to JP-A-10-269237, FXPAL Bar, Improving Proactive Information Systems, ACM IUI 2005, and Margin Notes, Margin Notes: Building a Contextually Aware Associative Memory, ACM IUI 2000).
Incidentally, the documents being retrieved results have their titles or summaries, their retrieval keyword, etc. displayed in juxtaposition with the Web page being browsed, or they are presented in a separate window or as a list which can be traced from the menu of an application.
With the related-art information recommendation system as stated above, the documents recommended by the system are determined on the basis of their degrees of similarities to the contents of the Web page being browsed by the user. The system, however, has had the problem that the extents to which the recommended documents are actually valuable for the user are unknown.